Smoking is the leading preventable cause of death and morbidity in the United States. The chief aim of this research is to demonstrate that particular kinds of smokers quit smoking more successfully with one treatment than another. Such a result would permit the optimal matching of smokers with cessation treatments and increase smoking treatment efficacy. The two smoking cessation treatments contrasted in this research are support counseling and skill training. These are frequently used treatments and are easy to implement. 560 smokers, stratified on their pretreatment affect, will be randomly assigned to one of four interventions: Skill training, Support counseling, combined Skill training and Support counseling, or an Attention/Placebo control condition. All subjects will receive the nicotine patch as an adjuvant. The three active- counseling treatments should produce significantly greater long-term abstinence than the control treatment. The most important predicted result is that subjects' pretreatment negative affect (mood) will interact with treatment type. Among subjects high in negative affect prior to treatment, Support counseling will produce better outcomes than any other treatment. Among subjects with little pretreatment negative affect, Skill training will produce better outcomes than any other treatment. Finally, this research will reveal the mechanisms by which the various treatments produce their beneficial effects, and materials will be developed that will promote the widespread use of the identified matching rules. Because quitting smoking has large health benefits, this research should reduce death and disease produced by smoking.